agentleFS
analyticsengineeringanalyticsengineering/ae-toolkit/skills/rename-model/SKILL.md

rename-model

Skill3 starsChanged 31 days agoMITView on GitHub

Coordinates name-only renames of analytical models, columns, relations, sources, metrics, semantic objects, and every repository reference while preserving the data contract. Use for atomic internal column renames and cross-layer reference updates; use migration-plan when behavior/schema changes or a published consumer interface needs compatibility and phased cutover.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. About 0 are common, 11 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not combine a rename with intentional grain, row, schema, or meaning changes.

    line 22
  2. Very specific:

    Read .ae/exercise.json when present and honor its mode and constraints.

    line 26
  3. Distinctive:

    Read the model, configuration, documentation, tests, macros, and project configuration that establish its logical name, physical relation, …

    line 33
  4. Distinctive:

    Before editing, deleting, or renaming any existing file, show the exact path map and reference diff, disclose the blast radius across …

    line 47
  5. Distinctive:

    After approval, apply only the approved path and reference changes.

    line 48
  6. Distinctive:

    Every actionable finding includes severity and confidence plus Evidence, Risk, Recommendation, and Verification.

    line 55
  7. Very specific:

    After an approved rename, return # Rename Result with: Paths created, changed, renamed, and removed.

    line 73
  8. Very specific:

    If the request also changes behavior or schema, return # Migration Required and name each changed contract element.

    line 75
  9. Distinctive:

    Search the full declared scope for the old logical name, relation name, aliases, and qualified forms.

    line 79
  10. Distinctive:

    Check that temporary aliases preserve columns, grain, permissions intent, and deprecation ownership.

    line 82
  11. Distinctive:

    Never claim external consumers are migrated without direct local evidence.

    line 84
  12. Distinctive:

    Never call MCP, execute dbt, query a warehouse, or run pipelines.

    line 89
  13. Distinctive:

    Never issue DDL, mutate production relations, or alter grants.

    line 90
  14. Distinctive:

    Never expose credentials, PII, or sensitive values found in configuration.

    line 91
  15. Very specific:

    Do not assume references use only ref, filenames define relation names, aliases are portable, external BI is searchable, consumers deploy …

    line 97

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md
<!-- Generated by scripts/build_targets.py from src/skills. Edit the source and rebuild; changes made here are overwritten. -->

Self-contained install. <skill-dir> is the directory holding this file: substitute its absolute path in the commands below and run them from the repository root. Read <skill-dir>/references/operating-rules.md before acting; it holds the toolkit's always-on rules, which this install cannot load on its own. This install has no safety hooks, so the approvals those rules describe rest on you and the agent rather than on an enforced gate.

Rename Model

Change a model's identity without silently changing the data it publishes or leaving consumers behind.

First decision: has this rename been approved?

An existing file changes only after the requester has seen the plan and approved it in this conversation. Until then the response is # Rename Plan and no file is edited, renamed, or deleted, however the request was phrased. # Rename Result belongs to the turn after that approval. A compatibility strategy, such as keeping the old name as an alias, is one of the options the plan presents for approval, not a way to make editing safe without it.

When to use

  • Rename a model, column, source, relation, metric, semantic object, or model file.
  • Inventory and update references coordinated across analytics artifacts.
  • Design a temporary alias or deprecation period for consumers that cannot move atomically.
  • Do not combine a rename with intentional grain, row, schema, or meaning changes.

Mode and context

Read .ae/exercise.json when present and honor its mode and constraints. Otherwise default silently to Work. In Learn mode, explain the dependency search progressively. In Assessment mode, evaluate the submitted rename without editing or completing it.

Obtain the old identity, intended new identity, project root, desired compatibility window when one is stated, and any known external consumers. Ask only whether a breaking rename is acceptable and on what timeline when repository evidence cannot answer it.

Procedure

  1. Establish repository conventions first. Read .ae/project.yml when present; otherwise run python3 "<skill-dir>/scripts/infer_conventions.py" <project-root> --json. Record naming, model paths, YAML layout, alias patterns, adapter, and transformation tool with evidence strength.
  2. Read the model, configuration, documentation, tests, macros, and project configuration that establish its logical name, physical relation, aliases, schemas, and database.
  3. Confirm the rename does not also change grain, rows, types, meanings, keys, materialization, or freshness. A column's old and new names may differ while its value and type contract remain identical. Route intentional behavior/schema changes and published column interfaces that require compatibility or phased consumer migration to migration-plan.
  4. Trace upstream and downstream identities using both logical and physical names. Search:
    • model, source, and column references; SQL identifiers and strings; macros; and package-qualified references;
    • YAML tests, contracts, documentation, groups, access, versions, and selectors;
    • exposures, metrics, semantic models, saved queries, LookML, and other BI artifacts;
    • orchestration, freshness checks, grants, monitoring, exports, seeds, snapshots, Python, and other scripts.
  5. List every searched directory and file type. Mark external systems or formats that local files cannot verify rather than treating them as clear.
  6. Classify consumers as atomic, compatibility-dependent, external/unknown, or generated. Never edit generated artifacts or distribution copies unless explicitly requested as source files.
  7. Choose a coordination strategy supported by repository and platform evidence:
    • atomic rename for a private/internal column or object when every local consumer can move together and no published interface is affected;
    • temporary compatibility alias/view when old and new names must coexist;
    • versioned or staged migration when ownership or release timing differs.
  8. Define ordering, ownership, deprecation date, monitoring signals, and removal criteria. Account for case-only renames and physical aliases on case-folding filesystems or warehouses.
  9. Before editing, deleting, or renaming any existing file, show the exact path map and reference diff, disclose the blast radius across models, tests, exposures, metrics, BI artifacts, schedules, grants, and external consumers, and obtain explicit approval. Stop before writing until approval arrives.
  10. After approval, apply only the approved path and reference changes. List each newly created path and each old path removed.
  11. Re-run the same static searches against old and new identities, then report unresolved consumers and compatibility obligations.

Evidence required

Cite the definition of the old identity, the convention supporting the new identity, and every discovered reference by file and exact token. Distinguish Observed repository consumers, Inferred physical dependencies, and Unknown external consumers.

Every actionable finding includes severity and confidence plus Evidence, Risk, Recommendation, and Verification.

Failure modes

  • Renaming only the SQL file or only the ref calls.
  • Renaming a column in SQL while leaving YAML tests, contracts, Semantic Layer dimensions, metrics, or BI fields on the old name.
  • Assuming the logical model name equals its physical relation name.
  • Missing string-based SQL, selectors, grants, monitors, exports, or BI references.
  • Mixing cleanup or schema changes into a supposedly name-only operation.
  • Dropping the old identity before compatibility consumers have moved.
  • Keeping an alias indefinitely without owner, deadline, or removal criteria.
  • Treating generated manifests or distribution copies as source-of-truth files.
  • Editing first and asking for approval afterward.

Output contract

Until approval has been given in this conversation, return # Rename Plan with: Old and new identities; Contract-preservation check; Convention evidence; Reference inventory; External unknowns; Compatibility strategy; Ordered path and reference diff; Blast radius; Verification plan; Approval needed.

After an approved rename, return # Rename Result with: Paths created, changed, renamed, and removed; References updated; Compatibility retained; Static verification; Unresolved external consumers; Follow-up commands not run.

If the request also changes behavior or schema, return # Migration Required and name each changed contract element.

Verification

  • Search the full declared scope for the old logical name, relation name, aliases, and qualified forms.
  • Confirm every new reference resolves to an existing local definition.
  • Confirm tests, contracts, documentation, selectors, exposures, and semantic objects use the intended identity.
  • Check that temporary aliases preserve columns, grain, permissions intent, and deprecation ownership.
  • Propose the smallest targeted compile, build, or consumer validation commands, but do not execute them.
  • Never claim external consumers are migrated without direct local evidence.

Safety

  • Local-only: use repository files, local static tools, and existing artifacts.
  • Never call MCP, execute dbt, query a warehouse, or run pipelines.
  • Never issue DDL, mutate production relations, or alter grants.
  • Never expose credentials, PII, or sensitive values found in configuration.
  • New files may be created without approval and every path must be reported.
  • Existing-file edits, deletions, and renames always require preview, blast-radius disclosure, and explicit approval immediately before the change.

What not to assume

Do not assume references use only ref, filenames define relation names, aliases are portable, external BI is searchable, consumers deploy atomically, grants follow a renamed relation, or a case-only rename behaves consistently across platforms.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.